Ferguson Family University Fellows Endowment

This summer, I had the invaluable opportunity to do a research fellowship. Throughout the project, I had great chances to advance my quantitative and analytical toolkit. I learned how to extract, clean, and harmonize massive datasets. I also developed more knowledge in econometric techniques such as structuring panel data and running Difference-in-Differences regressions using R and Python. But most importantly, beyond these technical skills, this experience taught me the importance of resilience.

During my fellowship, I focused on finding ways to optimize air sampling techniques. Volatile organic compounds (VOCs) include a diverse group of chemicals that easily evaporate into the air and can have various origins. Exposure to VOCs has been linked to serious health effects, making it a significant public health concern. This fellowship worked to advance the already used methods of evacuated canister sampling and low-cost sensors (LCS).

During my fellowship, I focused on the design and analysis of regular expressions (regex), which are symbolic notations used for defining search patterns in text. While widely used in programming and data validation, poorly written regex can introduce serious performance vulnerabilities, particularly Regular Expression Denial of Service (ReDoS) attacks. These occur when specific input patterns cause excessive backtracking and resource exhaustion.

During my fellowship, I developed a cross-platform physical therapy app to provide accessible, customizable solutions for athletes and individuals managing chronic injuries. I focused on learning Flutter and Dart to build the app, starting with a Minimum Viable Product (MVP) that included hard-coded routines, a simple user interface, and data handling capabilities. 

As artificial intelligence (AI) continues to advance, conversational AI systems like Chat GPT have emerged as powerful tools for natural language processing and human-computer interactions. My research aims to provide a thorough examination of the strengths and weaknesses of Chat GPT, while also delving into the pervasive issue of biases inherent in various AI models.

Using China’s national and regional data (i.e. Shanghai, Beijing, and Guangdong), this project studied how Covid-19 affects China’s housing market. Main findings included: